Abstract

This paper describes a case study about the creation and use of a prenatal clinical-information system developed at a Western New York Health Managed Care Organization (MCO). The authors identify the “conception-reality gap” for development and implementation of a healthcare information system (HCIS). Using a case study format, the authors identify the differences in assumptions that occur when developing a (HCIS) from an information technology (IT) versus a healthcare-provider perspective. The success of the program was related to the environment with respect to technical, social and organizational factors, especially the perceptions of key stakeholders.

Keywords: Socio-technical factors, social informatics, conception-reality gap, informatics, health care outcomes

Introduction

The use of information and computing technology in the managed care arena can reduce healthcare costs by controlling utilization of services and improving quality outcomes. This paper utilizes a case study approach to describe the development and use of a prenatal clinical information system, developed by a Western New York Managed Care Organization (MCO), to evaluate a database to manage the patient care information of a particular population In the United States, managed care occurs in an environment stressing capitalism, individualism, and the perception of unlimited resources -- accompanied by an emphasis on developing and implementing knowledge- and wisdom-based “best practices”. Effectively designed and appropriately implemented information technology in a managed-care setting can reduce healthcare costs through accurate tracking and controlling of services, thereby improving quality outcomes.

Nurses traditionally document patient information in a medical record using pen and paper. With today’s fast-paced environment nurses’ need more than traditional paper sources of information, like the electronic medical record, for decision-making and monitoring outcomes. “Nurses must have reliable, accessible, up-to-the-minute data to make appropriate health care decisions that lead to improved patient outcomes across the continuum of care (McDaniel, 1997).” Patient outcomes are measured in today’s healthcare environment acknowledging the increasing complexity of care.

Collaboration between healthcare providers and IT professionals is often fraught with cognitive, cultural, social and technical challenges. Coiera (2002) emphasizes the importance of social ‘factors’ on the development of technical solutions. The notion of a “conception-reality gap” (Heeks, Mundy, & Salazar, 1999) is used to illustrate the development and implementation of a healthcare information system (HCIS) for the prenatal program. A case study format will be used to describe the development of this application, and identify differences in assumptions between information technology (IT) and healthcare-provider perspectives.

Literature Review

Healthcare

Healthcare providers need information to guide patient health management while administrators focus on resource management. Each respective stakeholder group demands quick access to information, efficient provision of services and resources, and ultimately decreased costs. Complicating these needs in the United States are (1) the development of incompatible, proprietary applications; (2) a lack of standards; (3) a design-reality gap; and (4) communication complications. The effort to develop a viable HCIS is further complicated by the intersection of the above and (1) the highly politicized managed-care environment; (2) the potential of the Internet to influence healthcare; (3) the development of telehealth; (4) the electronic medical record (EMR); (5) the use of large-scale databases for the management and analysis of clinical outcomes; and (6) the Health Insurance Portability and Accountability Act (HIPAA) (Sackett & Erdley, 2002).

Spil & Stegwee (2001) noted that healthcare organizations are increasingly incorporating information technology into their change processes. This integration not only impacts organizational structure and individual responsibilities, but also organizational culture. Additionally, within healthcare organizations, information technology will have either beneficial or deleterious impacts, depending on the quality of the systems employed (Spil & Stegwee, 2001). However, information technology is in and of itself a neutral variable. The authors emphasize organizations with well-defined information structures integrate these structures as a result of socio-technical relationships among participants.

Coeira (2002), Schoop & Wastel (1999) and Vimlund & Timpka (2002) examine the importance of communication as a variable of success. The communication is not only among healthcare provides but between information technology persons and healthcare providers. Successful communication and collaboration between developers and users more often than naught will lead to successful application implementations.

Social/organizational

Millions of dollars have been spent to create a variety of HCIS that may or may not have proved useful in healthcare settings. Heeks, Mundy & Salazar (1999) detail reasons for the success or failure of HCISs via a conception-reality gap. This gap is comprised of three specific design-reality gaps identified as "(1) rationality-reality gaps - the formal, rational way in which many HCIS were conceived, which mismatched the behavioral realities of some healthcare organizations; (2) private-public reality gaps - application in private sector contexts of HCIS developed for the private sector, and (3) country-reality gaps - application in one country of a HCIS developed in a different country (p.1)." These authors also developed an acronym based on an assessment of Information, Technology, Process, Objectives and values, Staffing and skills, Management and structures, and Other resources: money and time (ITPOSMO) to delineate the specific design-reality components of an HCIS. Examination of the ITPOSMO dimensions suggests that a successful HCIS is able to ". . . match its environment in relation to technical, social and organizational factors . . . including the perceptions of key stakeholders (p.3)."

Berg (2001) looks at information system implementation in healthcare organizations with a focus on the concepts of success and failure. He noted some organizations might try to work through system troubles by pouring more resources into correcting perceived problems. Others instead may continue to stumble along with the system as is. “In the end, then, the question whether an implementation has been successful or not is socially negotiated (p.3).”

For Berg, components of a successful HCIS implementation include “effectiveness, efficiency, organizational attitudes and commitment, worker satisfaction, [and] patient satisfaction (p. 5).” Success of any system, then, becomes dynamic and requires evaluation over time by end-users, IT personnel, and administrators. He also notes "adequate user-involvement . . . is of paramount importance to foster ownership of the system by the future users, and to allow the implementation of systems that will actually match work processes -- current or future (p.8)." Additionally, van der Meijden, Tange, Troost, & Hasman, (2003) incorporate setting / environment and objectives as equally important determinants of successful implementation.

Technical

Kling (2000) explores the intersection of social informatics and information technology. Social informatics examines “. . . the design uses, and consequences of information technologies in ways that take into account their interaction with institutional and cultural context (p.1).” Information technology has been traditionally perceived as a tool, which is not a complete picture. "The standard Tool Model tends to both underestimate the costs and complexities of computerization, and overestimate the generalizability of applications from one setting or group of individuals to another ((Kling, 2000), p.29)." IT is increasingly viewed to be a part of a social-technical network, socially influenced and shaped. He also reports "a large body of research . . . which reports that the consequences of computerization depend upon the context in which systems are developed, implemented, and used (p. 5)." His socio-technical approach to HCIS design requires designers use a discovery process to ". . . effectively understand the relevant life worlds and work worlds of the people who will use their systems (p. 18)." In essence, this discovery process assists system designers to understand those features and tradeoffs most acceptable to the people who are most likely to use the system. This sort of approach incorporates user participation in system design teams as well as additional user participatory strategies.

Barnes (2001) utilized a grounded theory methodology to examine implementation of information systems in the United Kingdom. He identified numerous variables that impacted implementation from a contextual perspective. These included communication limits, environmental constraints (or user requests), organizational politics, structure, antecedent projects, and technical limitations. (Suomi, 2001) offers reasons for the delay in information technology adoption by healthcare organizations in spite of the information-intensive nature of healthcare. Of these reasons, the myopic control of information technology by IT personnel and the impositions of organizational structure are the most relevant to this paper.

Case Study

Background

In 1996 an MCO in Western New York developed a prenatal healthcare program to identify high-risk mothers. The target population for the program was all pregnant members in the organization’s largest health plan, which has approximately 569,000 members. On average, this member population has 4100 births per year. The purpose of the prenatal health program was to ensure the identified high-risk mothers received the necessary care to maximize the likelihood they would deliver healthy full-term babies.

The need for a prenatal program focused on high-risk pregnancies drew support from the region’s high incidence of infant mortality (8.8%), pre-maturity (11.3%), and low birth weight infants (7.7%) (Sackett, Pope, & Hendricks, 2001). Low birth weight (LBW) is defined as the delivery of an infant less than or equal to 2500 grams. Childbirth and LBW births in particular, comprise one of the highest costs of the MCO (over $200,000 can be spent for a single LBW infant). Overall, more than eighty million dollars is spent annually on prenatal care, making this the largest expenditure for this MCO.

To support this program, a software application was developed, implemented, and evaluated. The prenatal program software application integrated telephonic and IT as solutions for development and administration of the total program.

Conceptual Problems

When the original prenatal application was developed, there was minimal interaction between the developer and the end user. The developer who built the application used criteria developed from outcome reports. No choices were offered to the nurse, the application’s principle user, as to user friendliness and efficient program changes or updates. What was created was a technically correct but unwieldy application. The organizational structure required submission of formal change requests for any new reports or updates. For example:

• Early on, there was a data entry feature that did not allow the user to proceed to the next field unless all the data in the prior field were entered. This was a substantial limitation since maternal case management is telephonic and providers calling in would not necessarily have all of the data.
• A nurse could not override any of the preset application choices. The system was set to identify a woman as a high-risk pregnancy if any question on her history and physical report was positive. So, for example, if you had your tonsils out at age 8, the system would label you a high-risk pregnancy.
• The original system did not link with the MCO’s mainframe-based claims system. Therefore, paper reports were generated from several database systems. End users were required to manually compile the data into a final comprehensive report.

It became apparent the initial programmer was focused on technical issues. He used a rational model with assumptions of logic, formality, and objectivity, all visibly evident in the application. This model did not solicit a nursing perspective. In spite of the original developer’s extensive experience and reputation as a skilled programmer, the application he had developed for the program was difficult to use. To further compound the difficulties, once the program was completed, this programmer was assigned to other corporate software responsibilities.

This prenatal program was the first time a nurse and a programmer collaborated on a clinical project at this MCO. There was limited input from the nurse concerning (1) program design; (2) the user interface; (3) data field screen placement; (4) data field lockouts; and (5) the need for comprehensive report control and generation. After one week using the first version, it was apparent to the nurse project leader communication between her and the programmer was inadequate.

Subsequently, a junior programmer was assigned to maintain the application. Unfortunately, this person did not have strong development skills in Paradox®, which led to lengthy delays before application changes were completed. These delays increased the level of frustration among the nurses who used the application. The lack of understanding within the MCO’s IT department concerning necessary programming skills to successfully contribute to the application’s development resulted in increased user frustration and a further delay in achieving a viable HCIS. Consequently, this programmer too was replaced.

The next developer was talented but inexperienced. Hired immediately out of college he had a progressive perspective, complemented by an irreverent attitude and limited understanding of a formal business environment. This programmer built the second version of the prenatal application in Delphi® (a proprietary database application from Borland Software?™).

Up to and including this application version, data had to be transferred from the previous system to the new version. Sometimes this was met by auto-filling the new fields. Nurses perceived each new version as an opportunity to update and improve the application. As a result of updates, many of the date fields to be auto-filled were deleted or altered such that they no longer matched the data to fill them. Thus, the data from previous versions would not match the new fields, thereby creating complications generating outcome reports. This led to continual requests for report changes, which required adding new fields.

These communication difficulties arose because technical services at the MCO required specifications for a "finished product" application design from nurses, which the nurses were not equipped to provide. This let to a sense of frustration on the part of the programmers, who were used to receiving requests listing all specifications required for a finished product.

Further complicating application development, a communication lapse within the technical services department resulted in the application’s development using an application (Delphi) not supported by the IT department after the programmer departed. Thus, even though the nurses liked this second version of the program because of increased development input, it was eventually scrapped.

The experiences of the nurse who coordinated the project compelled her to insist she and the next programmer work closely together on the application. The goal of this close collaboration would be to match the application with the clinician’s needs rather than programmer preferences. The hiring of this programmer, who had prior experience working with nurses, improved the development process.

The third developer, with experience in Microsoft Visual Basic®, worked closely with the nurse leading the project. This developer rewrote the core program for the prenatal application using Visual Basic® 6.0. After working with the current programmer for over three months, the nurse leading the project and the programmer were able to agree on the design and terminology to be used to label data fields in the program, and the system’s special features were adapted.

The rewritten application version employs commonly accepted medical terminology to label data fields. This system also allows users to override preset data fields, and provides flexibility to move from screen to screen regardless of whether data fields are filled. Instead, an automatic reminder message appears requesting missing information be obtained. A monthly report is also generated to list missing information.

The greatly enhanced flexibility of the program stemmed from an eventual realization that "adequate user-involvement . .. is of paramount importance to foster ownership of the system by the future users, and to allow the implementation of systems that will actually match work processes -- current or future ((Berg, 2001), p.8)." Demonstrating the effectiveness of this principle, the improvements in the application contributed to the program’s recognition as a best practice by the New York State Department of Health.

Selected examples of the program:

• Entry and retrieval of case data: The application’s data entry fields and screens are laid out to follow the steps followed in the case management process. Each survey instrument has its own page, and additional pages show enrollment information, current status, and care notes. The application has its own database, but is also linked to the main member/provider database. This eliminates the need to store redundant information, while providing the most current data on members and providers. Linking to the mainframe allows the nurses to access and analyze existing clinical and demographic data.
• Transfer of data to and from ancillary services: Enrollment triggers the need to send educational gift packs and videos to members, reimbursement to providers, and other tasks.
• Report production: The application produces reports of requested variable by the case manager, such as patient age, low birth weight, smoking status, teen pregnancies, physician participation, and others.
• The ability to identify hospitalized and at-risk members. Clicking on an icon H brings up a list of all patients who have been hospitalized. In addition to patient names, all on-site (hospital) nursing notes are automatically transferred into the program. This provides the nurse with a complete review of all patients without having to contact the physician or hospital personnel.
• The ability to set the program to target any risk factor. For example, if a doctor indicates the patient is a smoker, the member automatically receives smoking cessation information.

Because of its success, this application will be employed on a broader scale, which will include multiple disease-management programs with the ability to make end-user customizations of data entry and reports.

Discussion

The prenatal application can be conceptualized as a set of interrelated decisions about technology and the organization of work. According to (Kling, 2000), this application typifies how organizations approach the development of computerized systems by simply installing and using new technology rather than ensuring the necessary socio-technical configurations. Throughout the original application development there was little collaboration between the nurses and the IT personnel as well as the sponsorship of the project, training on the use of the application, and control over access to information.

The struggle to create the prenatal program was illustrated by the difference between "‘where the HCIS wants to get us’ to ‘where the dominant stakeholders want the HCIS to get them’ (Sackett & Erdley, 2002)." What was needed was the use of reality-oriented techniques and applications, plus a participative approach that included all key stakeholders to close the gap between conception and reality. An appreciation of how to make computer information systems viable for end-users is required among skilled technical installers, trainers, and consultants proficient with advanced development applications. These professionals must be willing and able to learn from the end users themselves how to effectively integrate the applications and distribute into a particular social or work context.

In the MCO, nurses had minimal computer literacy and consequently were unaware of the application’s capacity and its limitations. This led many to avoid involvement in the design and creation process. The programmers involved had limited clinical knowledge. Prenatal terms, such as EDC (expected date of confinement) and gravida/para (number of pregnancies/number of live births), were as much a foreign language to them as programming and database terms were to the nurses. (Berg, 2001; Kling, 2000; Spil & Stegwee, 2001), each support the need for end users to be involved with IT personnel in the development and use of workable software applications. The software applications must effectively support end users’ work and communication requirements. (Berg, 2001) summarizes this best by acknowledging that ". . . users are generally very bad in speaking the language of ‘specifications,’ and in imagining what specific configuration of the technology they ‘need’ or what would work ‘best’ in actual work situations. Such judgment skills can only develop over time when users are taken on board in the development process early and systematically, and when careful attention is paid by those responsible for the implementation to the actual work processes that these users take as their starting point (p.9)." (Spil & Stegwee, 2001) lend additional support to this idea by acknowledging the importance of key stakeholders’ buy-in and IT personnel’s communication abilities as components of successful application development.

The prenatal program ultimately had the potential to significantly improve patient outcomes as well as the organization’s revenue. This was dependent on successful implementation of an application to support the data management requirement for the program defined by end-user satisfaction with the application. In this case study, as a result of the eventual high level of involvement of end users with the application’s development, the prevailing wisdom that there are more failures than successes with HCIS was avoided.

Lessons learned:

• Nurse end users should outline the purpose of the proposed application, including the major functional pieces, general inputs and outputs, and what constitutes “success.” This must be done in clinical terms and should include instructions for the developers as well. If they do not understand the language and processes being discussed, they must learn them.
• With the general process and objectives outlined, the developers must work with the end users performing the day-to-day tasks associated with the process the application will be intended to support. Observing completion of paperwork, how phone calls made, or how applications are interacted with is important to developing a digital plan of action. The developers should also encourage users to explain what they do and do not like about the current system.
• As the programming code is written, the nature of the processes should not be changed.
• Data storage should be in the form most “native” to the user’s needs. That is, store dates as dates and not as strings. Checkboxes are great space savers on a display, but may not correctly reflect the choices needed for many items.
• Reports should accurately reflect the criteria outlined by the user.

In short, the end user should drive the process, not the technology. The technology is a tool. Compromise is an important factor of this process. Sometimes a request is not practical and may require substantial labor for something that is not essential to the work process. Such requests should be pointed out to the user. On the other hand, the programmer should not resist a requested application feature merely because it’s difficult to accomplish.

(Barnes, 2001; Spil & Stegwee, 2001; Suomi, 2001) lend credence to the importance of socio-technical factors in defining a successful implementation. The necessity of information technology use in healthcare organizations is acknowledged along with the importance of socio-technical factors in application development, implementation and evaluation.

Summation/conclusions

We have presented explanations of the atypical HCIS success for the MCO’s prenatal program. Critical to the success of the application was the development environment with respect to technical, social, and organizational factors, especially as were manifested in the perceptions of key stakeholders (the nurse end user and the programmer). Successful conception-reality gap closure techniques for HCIS development, as defined by Heeks Mundy & Salazar (1999), support the production of reality-based applications. The techniques included:

• Extensive inclusion of the end user in the development of the HCIS.
• Customization of the software application to match the reality of the end users’ needs.
• Acknowledgement of the fact that the key stakeholders, in our example the nurse end user and the programmer, both function as critical change agents.
• Acknowledgement of the need for ongoing, incremental changes and improvements to the software application.
• An appreciation for the closure of the identified conception-reality gap so that the end users “take ownership” of the software application, allowing it to work in the “real world.”

In the case of the prenatal program’s computer application, along with countless others, its success or failure is not only a matter of technology, but of collaboration. (Vimarlund & Timpka, 2002) also support the importance of end-users participation in application development: “End-user participation in the design process can be the key positive influence on the quality of the service and, thereby, organizational effectiveness. Investments in broad design participation can, consequently, be a productive activity that transforms potential current income into future benefits. Today, though, the rigid technological perspective traditionally associated with computer science ‘training/education’ is evolving to include social factors, such as user involvement and collaborative communication, identified in this case study and by Vimarland and Timpka (2002). We outlined these factors and suggest strategies that may be applied in other MCOs to maximize the likelihood of success in the development of HCIS.

References

Barnes, S. J. (2001). Experiences in SIS Implementation in UK Healthcare. In R. A. Stegwee & T. A. M. Spil (Eds.), Strategies for Healthcare Information Systems (pp. 11-30). Hershey, PA: Idea Group Publishing.
Berg, M. (2001). Implementing information systems in health care organizations: Myths and challenges. (pp. 20). Erasmus University, Rotterdam, NL: Institute of Health Policy and Management.
Heeks, R., Mundy, D., & Salazar, A. (1999). Why health care information systems succeed or fail. Manchester, UK: University of Manchester, Precinct Centre, Instituted for Development Policy and Management.
Kling, R. (2000). Learning about information technologies and social change: The contribution of social informatics. The Information Society, 16(3).
McDaniel, A. M. (1997). Developing and testing a prototype patient care database. Computers in Nursing.
Sackett, K., & Erdley, W. S. (2002). History and Development of Health Care Informatics. In S. Englebardt & R. Nelson (Eds.), Information Technology in Health Care: An Interdisciplinary Approach (pp. 453-475). St. Louis, MO: Mosby.
Sackett, K., Pope, R., & Hendricks, C. (2001). Healthy pregnancy management: One company's example. The Case Manager, 12(1), 62-65.
Spil, T. A. M., & Stegwee, R. A. (2001). Section 1. Introduction: Strategies for Healthcare Information Systems. In R. A. Stegwee & T. A. M. Spil (Eds.), Strategies for Healthcare Information Systems (pp. 1-10). Hershey, PA: Idea Group Publishing.
Suomi, R. (2001). Streamlining Operations in Healthcare with ICT. In R. A. Stegwee & T. A. M. Spil (Eds.), Strategies for Healthcare Information Systems (pp. 31-44). Hershey, PA: Idea Group Publishing.
van der Meijden, M., Tange, H., Troost, J., & Hasman, A. (2003). Determinants of success of inpatient clinical information systems: A literature review. Journal of the American Informatics Association, 10(3), 235-243.
Vimarlund, V., & Timpka, T. (2002). Design participation as insurance: Risk-management and end-user participation in the development of information systems in healthcare organizations. Methods of Information in Medicine, 41, 76-80.

Authors’ Bios

Scott Erdley

Dr. Erdley has been a Clinical Assistant Professor of Nursing, and affiliated with the Patient Simulation Center, used by the Nurse Anesthesia and Acute Care NP Programs, since 2000. He received his Doctorate of Nursing Science from the School of Nursing, University at Buffalo, State University of New York. Dr. Erdley is also an alumnus of the School of Nursing having earned his B.S.N. and M.S. in 1989 and 1993, respectively. Recently he participated as a fellow in the Medical Informatics (MBL/NLM) Course held at the Marine Biological Laboratory, Woods Hole, MA (Spring 2002). Dr. Erdley’s interests and activities, with healthcare informatics, are currently in the areas of nursing information, information seeking, the use of high-fidelity simulation in education and health care informatics, and portable/personal digital devices. He manages several listservs including the Nrsing-l, the first list dedicated to nursing informatics. Dr. Erdley has taught classes about health care informatics as well as the intersection of technology and care in acute care environments to undergraduate and graduate students.

Kay Sackett

Dr. Sackett is currently an Assistant Professor of Nursing at the University at Buffalo, The State University of New York. Her academic credentials consist of a BSN, MEd and EdD in Health Education from Temple University in Philadelphia, PA; an MSN in Nursing Administration from Widener University in Chester, PA; and a Certificate in Online Teaching from UCLA, Los Angeles, CA.

Dr. Sackett became involved with technology and nursing in the mid 1990s. Nursing students were having difficulty passing the computerized NCLEX licensing examination. She attended a summer workshop on construction of computerized tests and a workshop on creating web pages in HTML. Her undergraduate student class assignments began to include computerized course examinations, building department, faculty and student web pages. A rudimentary “online” course was developed in nursing informatics.

Currently, Dr. Sackett teaches several web-enhanced courses for both undergraduate and graduate nursing students at the University at Buffalo School of Nursing. Active learning strategies are incorporated that use the WWW, videoconferencing, streaming video, web-based examinations, interactive PowerPoint presentations, and a myriad of other technology applications. She is also conducting research using the application of Perry’s Model of Moral and Intellectual Development in the College Years to student nurses’ adoption of technology applications.

While working toward the development of an education focused research program, Dr. Sackett began to examine the idea of data mining, databases and the use of the huge date repositories available to me at a local Managed Care Organization (MCO). She is working with the data that can be mined from the MCOs data rich repository of claims information. Dr. Sackett will be able to access information that can be used to determine the relationship of cost (return on investment and/or return on expenditure), and satisfaction to the use of nursing interventions [education focused] that are the core of disease management programs.

Rosemary K. Pope

Ms. Pope received her baccalaureate degree in nursing from D’Youville College in NY, is a Certified Case Manager and preceptor for the RN/BS students from the University at Buffalo (UB). She is a guest lecturer in the RN/BS nursing program and the Post Baccalaureate Program in Case Management at UB. Ms. Pope was awarded the “Best Practice in 2001” Program Award for the “Right Start” Prenatal Program that she coordinates by NY State. She has published articles on case management.